Chem.oBiol. Interactions, 9 (1974) 327-340 327 © Elsevier Scientific Publishing Company, Amsterdam - - Prinled in The Netherlands ISOTOPE EFFECTS IN O- AND N-DEMETHYLATIONS MEDIATED BY RAT LIVER MICROSOMES: AN APPLICATION OF DIRECT INSERTION ELECTRON IMPACT MASS SPECTROMETRY ALLAN B. FOSTER, MICHAI'L JARMAN, JOHN D. STI'V[:NS*, PETER THOMAS al~ll) JOHN H. WESTWOOD Chester Beatly Research Institute, Institute o/" ('anccr Rcs~'arch : Royal Cam'or Itospital. Fulham Road, Lon&'n Slt'3 6JB (Great BritainJ (Received Ma,'ch 19lh. 1974) (Revision received June 3rd, 1974) (Accepted July 2nd, 1974) SUMMARY Isotope effects of -2 have been found for the O-demethylafion of p-nitro- anisole, p-nxcthoxyacetanilide, an~, p-dinxethoxybenzene and the respective trideutero- methyl derivatives, when mediated by rat liver microsomes. The direct insertion mode o;'electron impact mass spectrometry (the advantages and limitations of wh.ich are discussed) was used together with conventional methods (observation of formaldehyde release, product analysis by spectrophotometry) to determine the isotope effects. Only the mass spectrometry n'telhod was applicabl~ for determining the isotope effect associated with the mono-O-demethylation of/,-tri- deuterotrtethoxyanisole and an t, nusually large value (10) was found. An insignificant isotope eiTect (> 1.05] was found for the mono-N-demelhyla- tton of I-to-carbamoylphenyl)-3,3-dimethy!triazene and its di-(trideuteromethyl) analogue. The protium and dctiterium forms had closely sin'tilar growth-inhit?itor? activities for the TLX5 lymphoma in mice. INTRODUCTION Drugs may be activated or deactivated by metabolism and one or more metab- olites may have activity compar~ ble to that of the initial drug t. Oxygenases are frequently involved in the initial ~netabolism of drugs especially in the liver. Alihough there are several oxygenase syste ~'.s-" each apparently activates molecular oxygen and * Visiting Scientist, Institute of Cant er Research, 1972--1973. Present address: Chemistry Depart- ment. University of New South Wal, s, Kensington (Australia), Since this paper was submitted, the use of direct-in~rtion chemical-ionisation mass spectrometry coupled with stable isotope dilution for the quantitation ofquinidine and lidocaine in human plasma has b~en described (W. A, GARLANI) W. F. TRAJER AND S. D. Nit.SON. Biomed, Mass Specrrom., I [1974) 124--129. Abbreviations: AR. Analar, DMSO, dimeth, t sulphoxide; El. electron impact.
`
`Apotex Ex. 1006
`
`Apotex v. Auspex
`IPR2021-01507
`
`

`

`328 ^. no FOSTER et aL releases "oxene" (the formal equivalent of atomic oxygen) "a in response to the pre- sence of an apprc ~,riate nucleophilic centre in the molecule (drug, carcinogen, or other exogenous chemical) adsorbed by the oxygenase complex. Oxene is isoelectronic with carbenes with which its reactions appear to be analogous s in relation to insertion into a-bonds (especially C-H and N-H to give tile corresponding hydroxy derivatives) or addition to ..-t-bonds (especially aromatic C:,,C bonds to give epoxides). A third type of reaction 4 involves lone pairs of electrons as in the conversion of tertiary amines into N-o,ddes. If oxene inserts directly into an R-H bond to give R-OH then an isotope effect, with a consequent slower reaction rate, should occur for the corresponding R-D compound. Most of the isotope effects* reported to date involve the conversion C-H(D)--+ C-OH and a magnitude of ,,,2, as, for example, in the demethylation of o-nitroanisole s (1). Recently, however, a value of 7 has been reported for the oxidative metabolism of cotinine 6 (2) and we have encountered a value of 10 for the demethylation of p-tfideuteromethoxyanisole (see below). The maximum theo- retical 7 isotope effect is 18. Noa i --~ ~o~h, o2 • HCHO 1 L.J - 2 (Cotinlne) Sign:ficant changes in pharmacological activity are associated with isotope effects of ,,-2 in, for example, N-trideu',eronormorphine s [the effect on the pharmaco- logical activity of an OCD3 group in codeine (which is activated by O-demethylation ~) apparently has not been investigated], 3'-deuterobutethal ~°, and 3',Y-dideutero- pentobarbilal t t. An isotope effect of 1.9 was observed for the oxidative deamination of (+)-2-deuteroamphetamine ;2 (see also VREE el al.ta). An understanding of the structural features of organic compounds which con- trol the magnitude of isotope effects might allow more rational consideration of the modification of drug activity by deuteration at the site of metabolic attack. We now report on a series of microsomally mediated O- and N-demethylation reactions selected in connexion with a preliminary evaluation of the direct insertion mode of El mass spectrometry for the determination of isotope effects since this mode has potential advantages (see DiscussioN). G.l.c.-mass spectrometry has been used ~'-:rhrouglloui this paper the isotope effect refers to the hydrogen-deuterium system.
`
`Apotex Ex. 1006
`
`

`

`STUDIES OF I.C~-l,'rOPE EFFECT BY MASS SPI!CTROMETRY 329 to study isotope effects as, for example, reflected by the urit~ary excretion of (-~-)- amphetamine, (+)-N-isopropylamphetamine and their deuterated analogues ad- ministered to man ~" The isotope effect involved in the demethylation ofo-nitroanisole by rabbit liver microsomes s was determined by UV absorption spectroscepy and by monitoring the release of l\~rmaldehyde: oxene insertion into a C--H bond ,~f the OMe group (I) affords an unstable hemiacetal of formaldehyde which decomposes into formaldehyde and o-nitrophenol. The Ik~rmer method can be used on~3 . when there is a significant difference in the UV spectra of the starting material az'd product and both methods require that the metabolism of the compound and it~ deuterated analog:ae be car- ried out separately. Mass spectrometry not only allows the metabolism of microgram ~;.l"tounts of a drug and its deuterated analogue (separately or in admixture) to be studied but is also potentially applicable to a wide variety of met~.bolic reactions. The application of mass spectrometry in the determination of isotope effects is limited to compounds whicll give molecular ions or appropriate fragm.,nt ons of reasonable intensity. With the advent of field ionization ~'~, chemical ionizati¢:n ~-~. and field desorption ~' mass spectrometry molecular ions can be obtained for a mt~ch wider range of compoands than is possible by the use of the El method and th.: scope of the technique for determining isotope effects has been greatly incre;tsed. MATERIALS ANI) Mt!'I'ItOI)S Preparation of rat liver l~dcrosomes Microsomes were isolated from the livers o(' male Wistar rats (6 weeks old, fed ad libitum) for which the drinking water contained 0.11~; phenobarbitonc for at least three days prior to killing. Tixe nucrosomes were prepared routinely by the calcium aggregation method ~v. The washed microsomal fraction was suspended in 0.1 M Tris-HCI buffer (pH 7.5) so tha~ microsomcs isolated from 1 g of liver were con- taincd in a final volume of I hal. Such preparations were usc:t immediately and found to be as active in their demethylating ability as a microsomal fraction prepared by more conventional means z s. Tills observation has been confirmed by CtNr! et al.~° and BAKER et al. "° for rat and mouse liver microsomes, respectively. Analytical methods Formaldehyde was determined by the procedure of NASH 2 t and p-nitrophenol by the method of McMataoN et al. z2. Mass spectrometric analysis of samples was carried out by the direct insertion technique using an AEI MS-12 spectrometer, an ionizing voltage of 70 eV, a trap current of I00 ,t,A, and an ion-source temperature of 60-100 °. For the determination of isotope ratios the chloroform solutions containing the products of metabolism and/or starting compounds with or without added reference standard (see RESUL'rS) were concentrated onto the direct insertion probe and each mass spectrum was scanned over a limited mass range which contained the peaks of the appropriate
`
`Apotex Ex. 1006
`
`

`

`330 A. !1. FOST~ et aL protium- and deuterium-containing molecular ions. Concordant ratios for each pair of peak heights (i.e. a variation of not greater than ~ ! % from the mean value ob~:ained from 5 scans) were obtained at a scan speed of 34 sec/decade in mass with a resolving power of 1000 providing that the total ion cuwrent monitor reading was first allowed to reach a steady value (of ref. 23). Synthesis of deuteronlethyl ethers Light petroleum refers to the fraction b.p. 40-60". Acetone was AR grade and was dried over potassium carbonate before use. Melting points were determined on a Kofler block and are corrected. The homogeneity of products was routinely moni- tored by t.l.c, using Kieselgel (7731, Merck) and conventional detection with sulphuric acid or Kieselgel GF2s,, (7730, Merck) and detection with UV light at 254 nm (Hanovia Chromatolite); Re values for individual compour~ds were similar for both adsorbents. p-Trideut~'romethoxyacetanilide. Sodium hydride (240 rag) was added to a cooled (0 °) solution of p-acetamidophenol (I.51 g) in N,N-dimethylformamide (10 mi) and the mixture was stirred for 20 min before the addition of trideuteromethyl iodide (2.84 g). After fur:her stirring and cooling for 30 min the solution was allowed to attain room temperature and then left for 24 h. Methanol was added and the mixture was concentrated to dryness to give the title compound (256 mg, 16 ,°,~), m.p. 127-129 ~ (from ethanol-light petroleum). Mass spectral data: m/e 168 (M +', 70°(,) 126 O/ ([M-CHzCO] +-, 68/o), and 108 ([126-CDa] ÷, 100~). p-Methoxyacetanilide, prepared by a similar procedure had m.p. 127-128". Mass spectral data: m/e 165 (M +', 80,°~), 123 ([M--CH2CO] +', 80,°/~), and 108 ([123-CH3] +, 100~o). p-Trideuteromethoxynitrobenzene. A mixture of p-nitrophenol (500 rag), silver oxide (2 g), N,N-dimethylformamide (4 mlL and trideuteromethyl iodide (0,75 ml) was stirred with the exclusion of light, for 1 h at 0" and then for 24 h at room tem- perature. The mixture was filtered and concentrated to a small volume. Chloroform was added and the )lution was washed twice with saturated aqueous sodium bicar- bonate, dried (MgSO,,), and concentrated to give the title compound m.p. 52-53" (from ethanol-light petroleum). Mass spectral data: m/e 156 (M +, 100'?,,,), 126 ([M-NO] +-, 18%), and 92 ([126-CD,~O] +, 40%~,). p-Methoxynitrobenzene (p-nitroanisole), prepared by a similar procedure, had m.p. 52-54". Mass spectral data: m/e 153 (M*', 100%,), 123 ([M-NO]*', 30~,,~), and 92 ([123-CH30] ÷, 40',',;i ). p-(Trideuterometh¢, r)anisole. A mixture of p-methoxyphenol (250 mg), tri- deuteromethyl iodide (0.2 ml), potassium carbonate (0.3 g), and acetone (3 ml) was stirred and boiled under reflux for 4.75 h. T.I.c. (ether-light petroleum, I : I) then showed a strong spot due to the title compound (Rr 0.65) and a weaker spot due to starting material (Re 0.45). After the addition of more trideuteromethyl iodide (0.1 ml) the mixture was boiled under reflux for 3 h, then cooled, and concentrated to dryness. A solution of the residue in benzene was twice extracted with M potassium h)droxide~ washed thrice with water, filtered, and concentrated. T.I.c. of the residue
`
`Apotex Ex. 1006
`
`

`

`STUDIES OF ISOI"OPE I~FFECr BY lklA~; SPECTROMETRY 331 showed only the compone,~t having RF 0.65. The residue was recrystallized from aqueous ethanol to give p-trideuteromethoxyanisole (108 nag), m.p. 55-56:'. A second crop (45 rag) was obtained from the filtrate. Mass spectral data: ode 141 (M +', 100~/,,), 126 ([M-CH~]', 50"i~), and 123 ([M-CI)3] +, 50°,3: there was virtually no peak at m/e 138. By using a similar procedure p-dimethoxybenzcne, m.p. 55-56", was obtained. p-Di-(trMeuteromethoxy)benzene. A mixture of hydroquinone (0.2 g), tri- deuteromethyl iodide (0.4 ml), potassium carbonate (0.25 g), and acetone (3 ml) was stirred and boiled under reflux for 6 h. The title compound, isolated using the above procedure, had m.p. 55-56 °. Mass spectral data: m/e 144 (M +', base peak): there were virtually no peaks at ,,1/e 141 and 138. p-Trideuterometho.~vphenol. A mixture of hydroquinonc (I g), trideuteromethyl iodide (0.2 ml), potassium carbonate (0.3 g), and acetone (3 ml) was stirred and boiled under reflux for 5 h. The residue left on ev;~poration of the solvent was par- titioned between water (20 ml) and benzene (20 ml). The benzene layer was washed with water (5 ml) and "he combined aqueous solutions were extracted with ethyl acetate (25 m[). The extract was washed with two small volumes of water, then added to the benzene solution and the mixture was concentrated. A solution obtained by treatment of the residue with ether (2 ml) and light petroleum (6 ml) was added to a column (height 23 cm) of Kieselgel (40 g, Merck, 7734, packed in etber-light petro- leum, I :4) followed by a second solution obtained by retreatment of the residue with ether (2 ml) and light petroleum (6 ml). The column was eluted with ethyl ether-light petroleum (I :4) and the fractionation was monitored by t.l.c. Com- bination and concentration of the appropriate (18-24) 10 ml fractions gave the title compound (180 rag) m.p. 54.5-55.5" (from benzene-light petroleum). Mass spectral data: m/e 127 (M +', 1000:,), and 109 ([M-CD3] +, 90)~,); there was virtually no peak at m/e 124. p-Methoxyphenol has m.p. 55-56". Mass spectral data: m/e 124 (M +', 100°:~,) and 109 ([M-CH3] +, 90~;,). I- ( o-CarboxyphtIo'l )-3,3-di- ( trhh, ut,,romethyl) triazene o-Carboxybenzenediazonium tetrafluoroborate -''~ (2.69 g) was added in small portions of finely ground solid, to a well stirred solution of di-(trideuteromethyl)- ammonium chloride (I g) and triethylamine (5.5 ml) in distilled water (I0 ml) and ethanol (5 ml) maintained at O. Tie mixture was stirred for I h, then acidified with acetic acid, and the resulting solid was extracted with ethyl acetate. The dried (Na.,SO.~) extract was concentrated and the residue was recryst~dlized from ethyl acetate after treatment with charcoal to give the title compound (1.25 g) as colour- less needles, m.p. 127 °. (Found: C, 54.23: N, 21.00%. C,~H.~D~,N30, requires C, 54.27: N, 21.107o.) I - (o- Carbamoylpheto'l)-3,3-di- (trhteuteromethyl) triazene A solution of but-2-yl chloroformate (i.3 ml) in tetrahydrofuran (25 ml) was added dropwise to a well stirred solution of 1-(o-carboxyphenyl)-3,3-di-(trideutero-
`
`Apotex Ex. 1006
`
`

`

`332 A. n. FOSlER et aL methyl)triazene (1.25 g) and triethylamine (3 ml) in tetrahydrofuran (25 ml~ main- tained at < 0 ° (ice-salt bath). After stirring for a further 30 rain anhydrous ammonia was bubbled into the mixture for 10 min. Evaporation of the solvent and recrystalliza- tion of the residue from ethyl acetate after treatment with charcoal gave the title compound (0.63 g) as colourless needles, m.p, 135-136". (Found: C, 54.37; N, 28.59% CgH6D6N,,O requires C, 54.54: N, 28.28'.'./). Mass spectral data: m/e 198 (M +', 25% of base peak at m/e 120); virtually no peak at m/e 192, I-(o-Carbamoylphenyl)-3,3.dimethyltriazene 2s prepared by a similar procedure had m.p. 135 °. Mass spectral data: role 192 (M +', 200{, of base peak at m/e 120) RESULTS O-Demethylation of p-nitroanisole and its trideuteromethyl mmlogue A solution of p-nitroanisole (100--1000 ttg) in 100 t~1 of DMSO was added to an NADPH generating system consisting of o-glucose 6-phosphate (30 ItM), NADP (I.5 ,uM), MgCI, (25 t~M), and D-glucose 6-phosphate dehydrogenase (4 units) in 0.1 M Tris-HCI buffer (pH 7.5, 7.9 ml). To this mixture a microsomal preparation (2 ml) was added and incubation was carried out at 37" for 30 rain with continuous shaking. The reaction was terminated by immersing the incubation flasks in a boiling water bath for 2 min. Precipitated protein was removed by centrifugation and the supernatant was twice extracted with chloroform (5 ml). The combined anU dried (Na2SO4) extracts were concentrated to dryness under reduced pressure. The residue was eluted with chloroform from a column (5 mm × 2.5 cm) of dry Kiese.lgel (7731, Merck). The methoxy derivatives were eluted slightly ahead of DMSO and the fractions containing UV absorbing material were combined and shown (t.l.c., chloro- form) to be free from DMSO which appeared as a diffuse spot with an RF value lower than that of the methoxy derivatives. DMSO interferes in the mass spectrometric analysis. For quantitative determination of the extent ol .... tabolism, an amount ofp-tri- deuteromethoxynitrobenzene equivalent to the initial quantity of p-nitroanisole was added as an internal standard to the incubation mixture immediately prior to the tetanination of the reaction. Comparison of the heights of the peaks for the molecular ions (raft, 153 for the OCH3 compound and 156 for the OCD~ analogue) enabled the amount of residual p-nitroanisole to be calculated. The metabolism of the deuterated compomtd was studied in a similar way using p-nitroanisole as the internal standard. Control mixtures of ~nown composition of p-nitroanisole and the trideuteromethyl analogue were subjected to the incubation and extraction procedure in the absence of co-factors. The results are presented in Fig. i as a double reciprocal plot of i/V vs. I/IS] where V is the rate of substrate conversion in/tmoles/30 min and [S] is the initial concentration of substrate; values obtained when a spectrophotometric method 2: was used to measure the amount of p-nitrophenol prc, duced are also included. After incubation for 30 min the extents of metabolisn, cf the OCH3 and OCD3 forms at a concentration of 7 • 10 -4 M were 32 and 17'}o, respectively. From Fig. I.
`
`Apotex Ex. 1006
`
`

`

`STUIM ,ES OF ISOTOPE EFFECT BY MASS SPECTROMJ'TRY 333 K,, values of 1.6.10 -'= M for p-nitroanisole and I.I • 10 -'= M for the trideutero- methyl analogue were obtained. From the extrapolated V values (OCH3 370, OCD 3 200 .,g of substrate demethylated during 30 rain) an isotope effect of 1.85 was cal- culated, MITOMA et al. ~ observed an isotope effect of ~2 and a lower K,. value for the OCD~ form in a study of tile demethylation of o-nitroanisole with rabbit liver microsonle,~. 007 ~" 0o3 NO2"~ °cHI O01 0 ~5 3030 6060 ~2 I 007 O05 003 00! A¢ ,H H lo 30 bo ,~o 90 1~'o ~ffo ~0 1 Isle' Fig. I. Effect of substrate concentration on tl-e metabolisn't of p-nitroanisole and trideuleromethyl analogue as determined by spectrophotometry (,,-', ,J~) and mass spectrometry (0. &); (~, O, p-nitro- anisole; .~,,, &, trideuteromethyl analogue. Fig. 2. Effect of substrate ,:oncentration on tile metabolism of p-methoxyacetanilide (Q) ,'rod p-tri- deuteromethoxyacelanilidc (,~) as determined by mass spectrometry. When an equin'tolar mixture ofp-nitroanisole and its trideuteromethyl analogue were incubated, then the relationship of the ratio (~,, as determined from the intensities of the molecular ions) and the extent of the metabolism (x, as determined by tile spectrophotometric method::) is given by tile equation I00 (a + b)-- 2ax -:: 1' [100 (a -!. b)-- 2bx] which may be solved for a/b which is the isotope effect. However. as the values of K,, for the deuterated and non-deuterated substrates may vary, the calculated isotope effect will vary slightly with variation in substrate concentrations. Only when the microsomal system is saturated with respect to each substrate will a true isotope effect be observed. The advantar:e of this me:hod, however, is that the metabolisms of the two substrates are directly comparable since both are present in the same incubate thereby ensuring identical conditions. O-Demethylation of p-methoxyacetani!ide Mixtures of p-methoxyacetanilide and p-trideuteromethoxyacetanilide were used to validate the mass spectrometric method as an analytical procedure. Thus, from
`
`Apotex Ex. 1006
`
`

`

`334 A.B. FOSTER et al. the ratios of the intensities of the molecular ions at m/e 165 and 168 for mixtures con- taining 20, 40, 60, and 80% of the OCD3 derivative, corresponding values of 20, 41, 61, and 83% were obtained. Experiments similar to those described for the metabolism of p-nitroanisole were performed on p-methoxyacetanilide and p-trideuteromethoxyacetanilide. The results are given as a double reciprocal plot in Fig. 2, The respective K,, values were 4.4.10 -'~ M and 3.6 • 10 -4 M, and the corresponding extrapolated V values (OCH.~ 260, OCD3 136 p g/30 rain) gave an isotope effect of 1.90. After incubation for 30 rain the extents of domethylation of p-methoxyacetanilide and p-trideuteromethoxy- acetanilide at a concentration of 7" 10 -4 M were 25 and 13 ,°,-o, respectively. O-Demeth)'lation of p-trideuteromethoxyanisole A solution of p-trideuteromethoxyanisole (100-I000 pg) in 0.1 ml of ethyl acetate was used but otherwise the incubation conditions were similar to those des- cribed for p-nitroanisole. An equivalent amount of p-dimethoxybenzene (p-methoxy- anisole) was added, as an internal standard, to the mixture after the incubation period. The incubation mixture was then extracted directly with ethyl acetate (2.5 ml) without prior boiling. Precipitated protein was removed by centrifugation and the ethyl acetate layer was dried (Na,SO4), filtered and evaporated under reduced pres- sure, and the residue was fractionated by t.l.c. [light petroleum (b.p. 60-80")--ethyl ether (4 : I)]. p-Trideuteromethoxyanisole (Rr 0.53) and the product (Re 0.15) of O-demethylation, namely a mixture of p-methoxyphenol and p-trideuteromethoxy- phenol, were located by their UV-absorbing properties. 1 i" 007 006 005 004 003 002 001 0005 o ¢o 2'0 3'o 4'0 s!o 6'0 I [s]'~-~ Fig. 3. Effect cf substrale concentration on tile metabolism of p-trideuteromethoxyanisole as deter- mined by mass spectrometry.
`
`Apotex Ex. 1006
`
`

`

`STUDIES OF ISOTOPE I'FFECT BY MASS SPECTROMETRY 335 The adsorbent containing each of these components was removed from the appropriate parts of the t.l.c, plate and extracted with methanol. The extracts were concentrated to dryness under reduced pressure and the residues analysed by mass spectrometry. The ratio of the peak intensities of Ihe molecular ions (m/e 124 and 127, re- spectively) of p-melhoxyphenol and p-trideuteromethoxyphenol is inversely propor- tional to the isotope effect" the obscrved ratio was I : 10. Analysis of the fraction containing p-trideuteromethoxyanisole and added p-methoxyanisole (molecular ions m/e 141 and 138, respectively) enabled the extent of metabolism to be measured. The double reciprocal plot of the total metabolism of p-trideuteromethoxyanisole (K,, 1.7 • !0 -3 M) is shown in Fig. 3 and Fig. 4 shows the extents of demethylation of the OCH~ and OCD3 groups. 80 o ,40 v o 2O 0 20 40 60 8'0 [s] ,~" Fig. 4. Effect of substrate concentration oll the rate of mono-O-demcthylation of the OCH.~ (~]) and OCD 3 groups (Q) of p.tridcuteromethoxyanisole. p-Dimethoxybenzene and p-di(trideuteromethoxy)benzene were separately incubated with microsomes and the results are given as a double reciprocal plot in Fig. 5. A conxp:~rison of results obtained using the mass spectrometric method with those obtained by estimating liberated formaldehyde-" showed close agreement (Fig. 5). An isotope effect of 2.1 was observed (respective extrapolated I / values, OCH.~ 143, OCD.~ 68 /~g/30 rain) and the respective K,, values were 1.8. 10 -3 M and 1.4. 10 -3 M. N-Demeth.t'lation of l.(o-carbamoylphenyl)-3,3-di-(trideuteromethyl)triazene A solution of the triazene (100-1000 ttg) in 100/,1 of acetone was used other- wise the incubation conditions were similar to those describect for p-nitroanisole, but using the appropriate protium or deuterium compounds as internal standards. The
`
`Apotex Ex. 1006
`
`

`

`336 ^. B. rOSTER et al. 0.(39 0(38 0.07 006 005 Q04 003 002 001 7 o ,!o ~io 3.'o 4'o ~'o ~o ,I IS] M'3 Fig. 5. Effect of substrate concentration on the mono-O-demelhylation of p-dimethoxybenzcne (C), O) and p-di(trideuteromethoxy)benzcne (A, A) as determined by mass spectrometry (O, A} and formaldehyde assay (9, A), 0.032 .1 1 ., I i 0 18'18 3636 7272 14,~I I IsiS-' Fig. 6. The N-demethylation of I-(o-carbamoylphenyl)-3,3-dimelhyltriazine and the di-~trideutero. methyl) analogue. Effect of substrate concentration on the rule of N,<lemethylation of the NMe2 compound (O, 0) and the N(CD3), analogue (A, A) as determined by formaldehyde release (0, A) and mass spectrometry (e, A),
`
`Apotex Ex. 1006
`
`

`

`STUDIES OF ISOTOPE ICFFECT BY MASS SPECTROMETRY 337 incubation mixtures were extracted with ethyl ;~cetate (2.5 ml) ~ithout prior boil- ing. Precipitated prbtein was removed by centrifugation and the ethyl acetate layer was dried (Na2SO.~h filtered, and concentrated under reduced pressure. The residue was eluted with methanol from a small column of Kieselgel (Merck, 7731). The triazene was located in the eluate fractions by t.l.c. (ethyl acetate-methanol, 5 : 1) and detection with UV light. The appropriate fractions were combined, concentrated to dryness under reduced pressure, and the residues were analysed by mass spectro- metry. Formaldeltyde prc, duction was measured 't directly on the incubaticn mixtures and the results ate presented in Fig. 6 as a double reciprocal plot. The data in Fig. 6 reveal a negligible isotope effect. Experiments performed on mixtures of the deuterium and hydrogen compounds also showed no isotope effect. I)ISCtlSSION Provided that the compounds studieu give either molecular ions (M -~) of reasonable intensity or suitably abundant fragment ions which retain the deuterium atoms present in th ~ molecular ion, then mass spectrometry is an analytical technique suitable for determining hydrogen-deuterium isotope effects in oxene insertion rent- lions. The OCH~ ~snd OCD3 compounds investigated herein gave molecular ions which were either t'le base peak or had intensities -.~:: 70~', of that of the base peak. Because of the closely similar volatility of the protium (e.g. R-OCHb) and deuterium forms (e.g. R-OCD~) the peaks of the respective molecular ions (Ma +, Mr, +) in the mass spectrum of :m equimolar mixture will usually have identical or closely similar intensities. Titus, the change in the ratio of peak intensities (Ma +, Mo +) will reflect the difference in rate of metabolism of the protium and deuterium forms. Although this ratio will be constant for a mixture of protium and deut~.rium forms of fixed composition the intensities of the peaks may vary as each sampl,~ is subjected to direct insertion El mass spectrometry because of the varying amount of vapourised material in the ion source. The extent of metabolism can be determined by measuring the Ma+/Mt~ + ratio before and after the addition of a known amount of the protium or deuterium form as an internal standard. Where the protium form alone is being metabolised then the deuterium form can be used as an internal standard and vice versa. A potentially important advantage of this procedure is that it is unnecessary to completely extract the protium and/or deuterium forms before and after the addi- tion of the internal standard to the microsomal system (or blood or urine) since selective losses are usually insignificant and the ratio Ma+/M~ + is independent of the amount of a mixture introduced into the ion source. Also, providing that the con- taminants do not yield io~s which give peaks at the positions of those of the molecular or appropriate fragment ions of the protium and deuterium forms, then there may not be a need for rigorou ~. p,:-ification. The quantitative determination of drugs and metabolites using stable isotope dilution and g.l.c.-mass spectrometry has recently been described 26. Although g.l.c.-
`
`Apotex Ex. 1006
`
`

`

`338 A.B. FOSTER eta/. mass spectrometry has been used to study isotope effects la not all compounds are amenable to g.l.c, and, in field desorption mass spectrometry, which is of particular value for polar compounds of very low volatility, coupling to a g.l.c, unit is not practicable. The declining concentrations ofp-nitroanisole and its trideuteromethyl analogue on separate incubation with rat liver microsomes in th.- presence of an NADPH generating system were determined by mass spectrometry and the appearance of p-nitrophenol which, apparently, is not metabolised further by a spectrophotome:ric procedure 2t. The results in Fig. I show good agreement between the two procedures. In all eases the rate of demethylation was linear during the incubation, and an isotope effect of i.85 was observed. A similar isotope effect (1.90) was observed for p-methoxy- acetanilide and p-trideuteromethoxyacetanilide (Fig. 2). The metabolism ofp-trideuteromethoxyanisole cannot be followed by spectro- photometry or by determining the released formaldehyde. It exemplifies a molecular situation which is the reverse of that present in the preceding experiments in that one starting compound gives two products. Mass spectrometry allowed the declining con- centration ofp-trideuteromethoxyanisole to be determined together with the ratio of the products, p-methoxyphenol and p-trideuteromethoxyphenol (Fig. 3). Fig. 4 shows the extent of demethylation of the OCHa and OCD3 groups and demonstrates the very large isotope effect (10). In seeking an explanation of this large isotope effect, the demethylation of p-methoxyphenol and p-trideuteromethoxyphenol was ir.vestigated. Under i ncuba tioi conditions where p-methoxyanisole underwent 50°,.o monodemethylation, the extert ofdemethylation ofp-trideuteromethoxyphenol was negligible and that ofp-methoxy- phenol was less than 3.°~ (BgAY et al. "-~ noted that rabbit liver slices did not cause significant demethylation of p-methoxyphenol) so that any error in determining the large isotope effect observed for p-trideuteromethoxyanisole due to further metabolism of the products is negligible. The data in Fig. 5 are for the separate incubation of p-dimethoxybenzene (p- methoxyanisole) and p-di-(trideuteromethoxy) benzene with microsomes for which an isotope effect of 2.1 was observed. That the large isotope effect associated with the demethylation ofp-trideutero- methoxyanisole is not due to a s0ecific property of microsomes but is an adsorption phenomenon is suggested by recent observations -'s on the chemical oxidation of alcohols (-CH2Oh -,- -CHO) under heterogeneous conditions. Whereas an isotope effect of 6. I was observed for the oxidation of 2-hydroxymethyl-2'-hydroxydideutero- methyldiphenyl with silver carbonate a value of 2 was found for an equimolar mixture of 2,2'-bisIhydroxymethyl)diphenyl and 2,2'-bis(hydroxydideuteromethyl)diphenyl. O M°- / N--N- X \ o CONI-I 2 M • - N(C~CH2C ~ 3 4 ~,
`
`Apotex Ex. 1006
`
`

`

`STUDIES OF ISOTOPE EFFECT BY MASS SPECTROMI'TRY 339 The marked growth inhibitoryactivityof 3,3-dinlethyltriazenes(R-N- ~ N-NMe2, where R is aromatic) for a wide range of transplanted tumours 2s may be due to the generation of alkyhtting metabolites consequent upon enzymatic N-demethylation. However, I-(o-carbamoylphenyl)-3,3-dinlethyhriazene (3) (which has high activity against the TLX5 lymphoma carried by CBA/LAC mice -''~) and its di(trideuteromethyl analogue were N-demethylated by nlicrosomes at closely similar rates (Fig. 6, K,, 6. I0 -'~ M, I" 330 !~g/30 n/in, i.~otope effect ~ 1.05). These data suggest that N-monodemethylation of 3 does not involve a direct oxene insertion into a C-H bond of a CH~ (or DC3) gcoup. If an N-hydroxynxcthyl derivative is the immediate precursor of the forn/aldehyde produced in the N-denlethylation then it must be formed by a rearrangenlent of tile oxene adduct by a process in which cleavage of the C-H (or C-D) bond is not rate determining. Oxygenases are known to convert tertiary amines into N-oxides -'9. The microsomally mediated conversion of cyclo- phosphamide (4) into the 4-hydroxy derivative 5 is also no subject to an isotope effect'-L As might be expected by tile absence of an isotope effect for the N-demethyl- ation, the 3,3-dimethyltriazene (3) and the 3,3-di-(trideuteromethyl) analogue showed 2s closely sinlilar growth inhibitory activity for the TLX5 lymphoma. It is of interest that dinlethylnitrosamine (Me:N-N: O), which is thought to be activated by N-dcmethylation causes an incidence (26,',i) of liver tumours in rats which is significantly higher (3";;) than that of the di-(trideuteromethyl) analogue 3°. ACKNOWLEDGEMENTS This investigation was supported by grants to tile Chester Beatty Research Institute (Institute of Cancer Research : Royal Cancer Hospital) from the Medical Research Council (G973/786/K3. The A,E.I. MS-12 spectrometer was purchased on a special grant (G969/189/C) f,'om the Medical Research Council. The work was carried during the tenure by (P.T.) of an Alexander